Wypych Handbook of Solvents

37DIN 53174-2, Solvents for paints and varnishes and similar coating materials; methods of analysis for solvent mixtures; gas chromatographic method, 1992.

38DIN 53175, Binders for paints, varnishes and similar coating materials; determination of the solidification point (titer) of fatty acids, 1991.

39DIN 53245, Solvents for paints and varnishes; alcohols; supply specification, further properties and methods of test, 1994.

40DIN 53246, Solvents for paints and varnishes - Acetic esters - Delivery specification, further requirements and methods of test, 1997.

41DIN 53247, Solvents for paints and varnishes - Ketones - Supply specifications, further requirements and methods of test, 1997.

42DIN 53248, Solvents for paints, varnishes and similar coating materials - Gum spirit of turpentine and wood turpentines - Requirements and methods of test, 1995.

43DIN 53249, Solvents for paints, varnishes and similar coating materials - Dipentene - Requirements and methods of test, 1995.

44DIN 55651, Solvents for paints and varnishes - Symbols, 1997.

45DIN 55681, Solvents; stability testing of trichloroethylene, 1985.

46DIN 55682, Solvents for paints and varnishes - Determination of solvents in water-thinnable coating materials - Gas chromatographic method, 1994.

47DIN 55682/A1, Solvents for paints and varnishes - Determination of solvents in water-thinnable coating materials - Gas chromatographic method; Amendment A1, 1998.

48DIN 55683, Solvents for paints and varnishes - Determination of solvents in coating materials containing organic solvents only - Gas chromatographic method, 1994.

49DIN 55685, Solvents for paints and varnishes; alcohols; gas chromatographic determination of the degree of purity, 1992.

50DIN 55686, Solvents for paints and varnishes; acetic esters; gas chromatographic determination of the degree of purity, 1992.

51DIN 55687, Solvents for paints and varnishes; ketones; gas chromatographic determination of the degree of purity, 1992.

52DIN 55688, Solvents for paints and varnishes - Ethylene glycol ethers - Gas chromatographic determination of the degree of purity, 1995.

53DIN 55689, Solvents for paints and varnishes - Propylene glycol ethers - Gas chromatographic determination of the degree of purity, 1995.

54DIN 55997, Solvents for paints and varnishes - Deionized water - Requirements and methods of test, 1998.

55DIN 55998, Solvents for paints and varnishes - Propylene glycol ethers - Supply specification, further requirements and methods of test, 1998.

56DIN 55999, Solvents for paints and varnishes, ethylene glycol ethers; supply specification, further properties and methods of test, 1994.

2

Fundamental Principles

Governing Solvents Use

2.1 SOLVENT EFFECTS ON CHEMICAL SYSTEMS

Estanislao Silla, Arturo Arnau and Iñaki TuñóN

Department of Physical Chemistry, University of Valencia, Burjassot (Valencia), Spain

2.1.1 HISTORICAL OUTLINE

According to a story, a little fish asked a big fish about the ocean, since he had heard it being talked about but did not know where it was. Whilst the little fish’s eyes turned bright and shiny full of surprise, the old fish told him that all that surrounded him was the ocean. This story illustrates in an eloquent way how difficult it is to get away from every day life, something of which the chemistry of solvents is not unaware.

The chemistry of living beings and that which we practice in laboratories and factories is generally a chemistry in solution, a solution which is generally aqueous. A daily routine such as this explains the difficulty which, throughout the history of chemistry, has been encountered in getting to know the effects of the solvent in chemical transformations, something which was not achieved in a precise way until well into the XX century. It was necessary to wait for the development of experimental techniques in vacuo to be able to separate the solvent and to compare the chemical processes in the presence and in the absence of this, with the purpose of getting to know its role in the chemical transformations which occur in its midst. But we ought to start from the beginning.

Although essential for the later cultural development, Greek philosophy was basically a work of the imagination, removed from experimentation, and something more than meditation is needed to reach an approach on what happens in a process of dissolution. However, in those remote times, any chemically active liquid was included under the name of “divine water”, bearing in mind that the term “water” was used to refer to anything liquid or dissolved.1

Parallel with the fanciful search for the philosopher’s stone, the alchemists toiled away on another impossible search, that of a universal solvent which some called “alkahest” and others referred to as “menstruum universale”, which term was used by the very Paracelsus (1493-1541), which gives an idea of the importance given to solvents during that dark and obscurantist period. Even though the “menstruum universale” proved just as elusive as the philosopher’s stone, all the work carried out by the alchemists in search of these

illusionary materials opened the way to improving the work in the laboratory, the development of new methods, the discovery of compounds and the utilization of novel solvents. One of the tangible results of all that alchemistry work was the discovery of one of the first experimental rules of chemistry: “similia similibus solvuntor”, which reminds us of the compatibility in solution of those substances of similar nature.

Even so, the alchemistry only touched lightly on the subject of the role played by the solvent, with so many conceptual gulfs in those pre-scientific times in which the terms dissolution and solution referred to any process which led to a liquid product, without making any distinction between the fusion of a substance - such as the transformation of ice into liquid water -, mere physical dissolution - such as that of a sweetener in water - or the dissolution which takes place with a chemical transformation - such as could be the dissolution of a metal in an acid. This misdirected vision of the dissolution process led the alchemists down equally erroneous collateral paths which were prolonged in time. Some examples are worth quoting: Hermann Boerhaave (1688-1738) thought that dissolution and chemical reaction constituted the same reality; the solvent, (menstruum), habitually a liquid, he considered to be formed by diminutive particles moving around amongst those of the solute, leaving the interactions of these particles dependent on the mutual affinities of both substances.2 This paved the way for Boerhaave to introduce the term affinity in a such a way as was conserved throughout the whole of the following century.3 This approach also enabled Boerhaave to conclude that combustion was accompanied by an increase of weight due to the capturing of “particles” of fire, which he considered to be provided with weight by the substance which was burned. This explanation, supported by the well known Boyle, eased the way to considering that fire, heat and light were material substances until when, in the XIX century, the modern concept of energy put things in their place.4

Even Bertollet (1748-1822) saw no difference between a dissolution and a chemical reaction, which prevented him from reaching the law of definite proportions. It was Proust, an experimenter who was more exacting and capable of differentiating between chemical reaction and dissolution, who made his opinion prevail:

“The dissolution of ammonia in water is not the same as that of hydrogen in azote (nitrogen), which gives rise to ammonia”5

There were also alchemists who defended the idea that the substances lost their nature when dissolved. Van Helmont (1577-1644) was one of the first to oppose this mistaken idea by defending that the substance dissolved remains in the solution in aqueous form, it being possible to recover it later. Later, the theories of osmotic pressure of van´t Hoff (1852-1911) and that of electrolytic dissociation of Arrhenius (1859-1927) took this approach even further.

Until almost the end of the XIX century the effects of the solvent on the different chemical processes did not become the object of systematic study by the experimenters. The effect of the solvent was assumed, without reaching the point of awakening the interest of the chemists. However, some chemists of the XIX century were soon capable of unraveling the role played by some solvents by carrying out experiments on different solvents, classified according to their physical properties; in this way the influence of the solvent both on chemical equilibrium and on the rate of reaction was brought to light. Thus, in 1862, Berthelot and Saint-Gilles, in their studies on the esterification of acetic acid with ethanol,

discovered that some solvents, which do not participate in the chemical reaction, are capable of slowing down the process.6 In 1890, Menschutkin, in a now classical study on the reaction of the trialkylamines with haloalcans in 23 solvents, made it clear how the choice of one or the other could substantially affect the reaction rate.7 It was also Menschutkin who discovered that, in reactions between liquids, one of the reactants could constitute a solvent inadvisable for that reaction. Thus, in the reaction between aniline and acetic acid to produce acetanilide, it is better to use an excess of acetic acid than an excess of aniline, since the latter is a solvent which is not very favorable to this reaction.

The fruits of these experiments with series of solvents were the first rules regarding the participation of the solvent, such as those discovered by Hughes and Ingold for the rate of the nucleophilic reactions.8 Utilizing a simple electrostatic model of the solute - solvent interactions, Hughes and Ingold concluded that the state of transition is more polar than the initial state, that an increase of the polarity of the solvent will stabilize the state of transition with respect to the initial state, thus leading to an increase in the reaction rate. If, on the contrary, the state of transition is less polar, then the increase of the polarity of the solvent will lead to a decrease of the velocity of the process. The rules of Hughes-Ingold for the nucleophilic aliphatic reactions are summarized in Table 2.1.1.

Table 2.1.1. Rules of Hughes-Ingold on the effect of the increase of the polarity of the solvent on the rate of nucleophilic aliphatic reactions

In 1896 the first results about the role of the solvent on chemical equilibria were obtained, coinciding with the discovery of the keto-enolic tautomerism.9 Claisen identified the medium as one of the factors which, together with the temperature and the substituents, proved to be decisive in this equilibrium. Soon systematic studies began to be done on the effect of the solvent in the tautomeric equilibria. Wislicenus studied the keto-enolic equilibrium of ethylformylphenylacetate in eight solvents, concluding that the final proportion between the keto form and the enol form depended on the polarity of the solvent.10 This effect of the solvent also revealed itself in other types of equilibria: acid-base, conformational, those of isomerization and of electronic transfer. The acid-base equilibrium is of particular interest. The relative scales of basicity and acidity of different organic compounds and homologous families were established on the basis of measurements carried out in solution, fundamentally aqueous. These scales permitted establishing hypotheses regarding the effect of the substituents on the acidic and basic centers, but without being capable of separating this from the effect of the solvent. Thus, the scale obtained in solution for the acidity of

the α-substituted methyl alcohols [(CH3)3COH > (CH3)2CHOH > CH3CH2OH > CH3OH]11 came into conflict with the conclusions extracted from the measurements of movements by NMR.12 The irregular order in the basicity of the methyl amines in aqueous solution also proved to be confusing [NH3 < CH3NH2 < (CH3)2NH > (CH3)3N],13 since it did not match any of the existing models on the effects of the substituents. These conflicts were only resolved when the scales of acidity-basicity were established in the gas phase. On carrying out the abstraction of the solvent an exact understanding began to be had of the real role it played.

The great technological development which arrived with the XIX century has brought us a set of techniques capable of giving accurate values in the study of chemical processes in the gas phase. The methods most widely used for these studies are:

•The High Pressure Mass Spectrometry, which uses a beam of electron pulses14

•The Ion Cyclotron Resonance and its corresponding Fourier Transform (FT-ICR)15

•The Chemical Ionization Mass Spectrometry, in which the analysis is made of the kinetic energy of the ions, after generating them by collisions16

•The techniques of Flowing Afterglow, where the flow of gases is submitted to

ionization by electron bombardment17-19

All of these techniques give absolute values with an accuracy of ±(2-4) Kcal/mol and of ±0.2 Kcal/mol for the relative values.20

During the last decades of XX century the importance has also been made clear of the effects of the solvent in the behavior of the biomacromolecules. To give an example, the influence of the solvent over the proteins is made evident not only by its effects on the structure and the thermodynamics, but also on the dynamics of these, both at local as well as at global level.21 In the same way, the effect of the medium proves to be indispensable in explaining a large variety of biological processes, such us the rate of interchange of oxygen in myoglobin.22 Therefore, the actual state of development of chemistry, as much in its experimental aspect as in its theoretical one, allows us to identify and analyze the influence of the solvent on processes increasingly more complex, leaving the subject open for new challenges and investigating the scientific necessity of creating models with which to interpret such a wide range of phenomena as this. The little fish became aware of the ocean and began explorations.

2.1.2 CLASSIFICATION OF SOLUTE-SOLVENT INTERACTIONS

Fixing the limits of the different interactions between the solute and the solvent which envelopes it is not a trivial task. In the first place, the liquid state, which is predominant in the majority of the solutions in use, is more difficult to comprehend than the solid state (which has its constitutive particles, atoms, molecules or ions, in fixed positions) or the gaseous state (in which the interactions between the constitutive particles are not so intense). Moreover, the solute-solvent interactions, which, as has already been pointed out, generally happen in the liquid phase, are half way between the predominant interactions in the solid phase and those which happen in the gas phase, too weak to be likened with the physics of the solid state but too strong to fit in with the kinetic theory of gases. In the second place, dissecting the solute-solvent interaction into different sub-interactions only serves to give us an approximate idea of the reality and we should not forget that, in the solute-solvent interaction, the all is not the sum of the parts. In the third place, the world of the solvents is very varied from those which have a very severe internal structure, as in the case of water, to those

whose molecules interact superficially, as in the case of the hydrocarbons. At all events, there is no alternative to meeting the challenge face to face.

If we mix a solute and a solvent, both being constituted by chemically saturated molecules, their molecules attract one another as they approach one another. This interaction can only be electrical in its nature, given that other known interactions are much more intense and of much shorter range of action (such as those which can be explained by means of nuclear forces) or much lighter and of longer range of action (such as the gravitational force). These intermolecular forces usually also receive the name of van der Waals forces, from the fact that it was this Dutch physicist, Johannes D. van der Waals (1837-1923), who recognized them as being the cause of the non-perfect behavior of the real gases, in a period in which the modern concept of the molecule still had to be consolidated. The intermolecular forces not only permit the interactions between solutes and solvents to be explained but also determine the properties of gases, liquids and solids; they are essential in the chemical transformations and are responsible for organizing the structure of biological molecules.

The analysis of solute-solvent interactions is usually based on the following partition scheme:

where i stands for the solute and j for the solvent.This approach can be maintained while the identities of the solute and solvent molecules are preserved. In some special cases (see below in specific interactions) it will be necessary to include some solvent molecules in the solute definition.

The first term in the above expression is the energy change of the solute due to the electronic and nuclear distortion induced by the solvent molecule and is usually given the name solute polarization. Eij is the interaction energy between the solute and solvent molecules. The last term is the energy difference between the solvent after and before the introduction of the solute. This term reflects the changes induced by the solute on the solvent structure. It is usually called cavitation energy in the framework of continuum solvent models and hydrophobic interaction when analyzing the solvation of nonpolar molecules.

The calculation of the three energy terms needs analytical expressions for the different energy contributions but also requires knowledge of solvent molecules distribution around the solute which in turn depends on the balance between the potential and the kinetic energy of the molecules. This distribution can be obtained from diffraction experiments or more usually is calculated by means of different solvent modelling. In this section we will comment on the expression for evaluating the energy contributions. The first two terms in equation [2.1.1] can be considered together by means of the following energy partition :

Analytical expressions for the three terms (electrostatic, polarization and disper- sion-repulsion energies) are obtained from the intermolecular interactions theory.

2.1.2.1 Electrostatic

The electrostatic contribution arises from the interaction of the unpolarized charge distribution of the molecules. This interaction can be analyzed using a multipolar expansion of the charge distribution of the interacting subsystems which usually is cut off in the first term

which is different from zero. If both the solute and the solvent are considered to be formed by neutral polar molecules (with a permanent dipolar moment different from zero), due to an asymmetric distribution of its charges, the electric interaction of the type dipole-dipole will normally be the most important term in the electrostatic interaction. The intensity of this interaction will depend on the relative orientation of the dipoles. If the molecular rotation is not restricted, we must consider the weighted average over different orientations

where:

kBoltzmann constant

εdielectric constant

Tabsolute temperature

rintermolecular distance

The most stable orientation is the antiparallel, except in the case that the molecules in play are very voluminous. Two dipoles in rapid thermal movement will be orientated sometimes in a way such that they are attracted and at other times in a way that they are repelled. On the average, the net energy turns out to be attractive. It also has to be borne in mind that the thermal energy of the molecules is a serious obstacle for the dipoles to be oriented in an optimum manner. The average potential energy of the di- pole-dipole interaction, or of orientation, is, therefore, very dependent on the temperature.

In the event that one of the species involved were not neutral (for example an anionic or cationic solute) the predominant term in the series which gives the electrostatic

interaction will be the ion-dipole which is given by the expression:

2.1.2.2 Polarization

If we dissolve a polar substance in a nonpolar solvent, the molecular dipoles of the solute are capable of distorting the electronic clouds of the solvent molecules inducing the appearance in these of new dipoles. The dipoles of solute and those induced will line up and will be attracted and the energy of this interaction (also called interaction of polarization or induction) is:

rintermolecular distance

In a similar way, the dissolution of an ionic substance in a nonpolar solvent also will occur with the induction of the dipoles in the molecules of the solvent by the solute ions.

These equations make reference to the interactions between two molecules. Because the polarization energy (of the solute or of the solvent) is not pairwise additive magnitude, the consideration of a third molecule should be carried out simultaneously, it being impossible to decompose the interaction of the three bodies in a sum of the interactions of two bodies. The interactions between molecules in solution are different from those which take place between isolated molecules. For this reason, the dipolar moment of a molecule may vary considerably from the gas phase to the solution, and will depend in a complicated fashion on the interactions which may take place between the molecule of solute and its specific surroundings of molecules of solvent.

2.1.2.3 Dispersion

Even when solvent and solute are constituted by nonpolar molecules, there is interaction between them. It was F. London who was first to face up to this problem, for which reason these forces are known as London’s forces, but also as dispersion forces, charge-fluctua- tions forces or electrodynamic forces. Their origin is as follows: when we say that a substance is nonpolar we are indicating that the distribution of the charges of its molecules is symmetrical throughout a wide average time span. But, without doubt, in an interval of time sufficiently restricted the molecular movements generate displacements of their charges which break that symmetry giving birth to instantaneous dipoles. Since the orientation of the dipolar moment vector is varying constantly due to the molecular movement, the average dipolar moment is zero, which does not prevent the existence of these interactions between momentary dipoles. Starting with two instantaneous dipoles, these will be oriented to reach a disposition which will favor them energetically. The energy of this dispersion interaction can be given, to a first approximation, by:

rintermolecular distance

From equation [2.1.6] it becomes evident that dispersion is an interaction which is more noticeable the greater the volume of molecules involved. The dispersion forces are often more intense than the electrostatic forces and, in any case, are universal for all the atoms and molecules, given that they are not seen to be subjected to the requirement that permanent dipoles should exist beforehand. These forces are responsible for the aggregation of the substances which possess neither free charges nor permanent dipoles, and are also the

very electronegative elements (e.g., F, Cl, O and N), which is the case with water. The hydrogen bond can be either intermolecular (e.g., H2O) or intramolecular (e.g., DNA). The protagonism of hydrogen is due to its small size and its tendency to become positively polarized, specifically to the elevated density of the charge which accumulates on the mentioned compounds. In this way, hydrogen is capable, such as in the case of water, of being doubly bonded: on the one hand it is united covalently to an atom of oxygen belonging to its molecule and, on the other, it electrostatically attracts another atom of oxygen belonging to another molecule, so strengthening the attractions between molecules. In this way, each atom of oxygen of a molecule of water can take part in four links with four more molecules of water, two of these links being through the hydrogen atoms covalently united to it and the other two links through hydrogen bonds thanks to the two pairs of solitary electrons which it